A Flexible and Dynamic CSP Solver for Web Service Composition in the Semantic Web Environment

As the Web service composition requests planning with scheduling with many parameters, the composition by the planner alone has the limitations that apply to a more general and intelligent Web service composition. The research [1] suggests a combined architecture of Hierarchical Task Network (HTN) planning and Constraint Satisfaction Problem (CSP) to solve the above limitation. In this paper, we focus on the CSP system on the semantic Web for the combined service composer and its applied solver, which gives the final solution set satisfying the constraints. The CSP solver, which is a part of the combined architecture, is extended by applying the semantic Web concept so that it can automatically solve a given problem more flexibly. CSP ontology is designed and implemented to work on the extension and this integrates the CSP and semantic Web concepts in the conceptual layer. The extended framework has advantages, such as generality, extendibility, dynamic configurability and parallelism due to the application of the semantic Web concept and a distributed CSP.